Process-based model predicts seasonal variation in eDNA transport - a case study on Eurasian beavers in a small river

Abstract Robust methods to monitor species distributions are vital to ensuring successful conservation strategies, particularly in the case of conservation translocations. Environmental DNA (eDNA) from water samples is a cost-effective method to monitor species distributions without physical capture or disturbance. However, eDNA is vulnerable to long-distance transport depending on the hydrological and environmental characteristics which can lead to spatially false positives and ultimately inaccurate species distributions. Recently, the development of particle transport models has allowed researchers to integrate hydrological and environmental variables to predict how far eDNA will transport from a source point. Here, eDNA samples (n=218) were collected and quantified using digital PCR (dPCR) to study monthly changes in Eurasian beaver ( Castor fiber ) eDNA concentrations downstream of an enclosure which contained 4 - 5 beavers located in Scotland. The shortest eDNA transport distances (< 2 km) were observed in the summer which correlated with the lowest flows and highest temperatures. In contrast, throughout the winter eDNA was consistently detected up to 5.8 km downstream correlating with the highest discharge and lowest temperature. The eDNA transport model reliably reproduced the decrease in eDNA concentrations downstream of the enclosure, however there were challenges surrounding stream-specific decay rates following a confluence. To study localised species distributions, samples should be collected during summer low flow conditions. Conversely, to maximise species detections sampling should be conducted in winter which had the longest eDNA transport and highest detectability.